The purpose of this study was to measure the specificity of the 21-Item Test for detecting malingering in an elderly sample. The 21-Item Test was administered to 20 community-dwelling elderly individuals with means for age, education, and Short Portable Mental Status Questionnaire (SPMSQ) error scores of 76.25 years (SD = 7.19), 14.30 years (SD = 3.89), and 0.30 (SD = 4.70). None of these healthy elderly participants were cognitively impaired. A second group of 53 nursing home residents had means for age, education, and SPMSQ error scores of 82.79 years (SD = 8.56), 11.14 years (SD = 2.46), and 3.17 (SD = 2.46). Approximately 60% of the nursing home participants had cognitive impairment based on the SPMSQ. The cut offs provided by Iverson (1998 , 21 Item Test Research Manual) for interpretation of the forced-choice component of the 21-Item Test were insensitive to age and relatively insensitive to cognitive deficits. Specificity was 100% since no elderly participant from either group scored in the range indicative of sub-optimal effort or biased responding.

No abstract